Research of Study Early-Warning Application Based on Association Mining

Because of its important application value in almost every region, early-warning has received extensive concern. This paper puts forward a study early-warning mechanism based on association rules. It uses an Apriori mining algorithm with some corresponding restrictions to dig out the latent school record association rules from former students´ scores which are viewed as a history resource. Then these rules will be used to match up the data sets that need to monitor. Once a record is matched to one of these rules, the student of this record will receive an early-warning. This kind of early-warning mechanism changes the situation that problems with study can only be detected after knowing the scores. It has a predictable and forward-looking capability and has been proved to obtain a good early-warning effect during actual verification.